FPGA-Based Degradation and Reliability Monitor for Underground Cables

Abstract

The online Remaining Useful Life (RUL) estimation of underground cables and their reliability analysis requires obtaining the cable failure time probability distribution. Monte Carlo (MC) simulations of complex thermal heating and electro-thermal degradation models can be employed for this analysis, but uncertainties need to be considered in the simulations, to produce accurate RUL expectation values and confidence margins for the results. The process requires performing large simulation sets, based on past temperature or load measurements and future load predictions. Field Programmable Gate Arrays (FPGAs) permit accelerating simulations for live analysis, but the thermal models involved are complex to be directly implemented in hardware logic. A new standalone FPGA architecture has been proposed for the fast and on-site degradation and reliability analysis of underground cables, based on MC simulation, and the effect of load uncertainties on the predicted cable End Of Life (EOL) has been analyzed from the results.